I'm no Elsevier fan, but I think that number is a bit over the top. Most Universities have site-wide access to pretty much any paper you can get your hands on. I have a really hard time believing that 10-20% of research funding feeds those journal databases.

You are correct: I am even less of an Elsevier fan that you, I'll warrant, but even their absurd prices "only" constitute on the order of 2% of the total cost of research.

But of course this makes it all the more ridiculous that we allow them to lock up the results of research. The issue here isn't just that Open Access is more economically efficient (although it is, by a factor of 3-5). The real issue is that there's so much can't do at any price with the current system or paywalls. If the Web as a whole were paywalled as academic papers are, we'd have no Google for example, because it wouldn't be possible to spider. We can't even SAY what the advantages to ubiquitous open access will be yet -- they haven't been invented, any more than spidering web-search engines were invented before the Web.

It's not just about the costs to Elsevier. It's about the costs of keeping information a secret that could have a much greater financial and societal impact if it were unleashed.

This is the age of the individual contributor. Open Source has taught us all about the power of appealing to and capturing the output of people working at home, in small groups, at small startups, etc.

So this isn't so much about recapturing the piece of a fixed pie that Elsevier takes. This is about making an incredibly larger pie by opening the information up to a wider audience and allowing us to compound the benefits in a much larger ecosystem.

Exactly! The real issue here is opportunity cost. Yes, we could have a much cheaper academic publishing system if we did it without the paywalls and the profiteering corporations. But that savings are as nothing compared with all the new application avenues that will open up when research is freely available.

I doubt the 10-20% number as well, but note that that "site-wide access" is essentially paid for out of research funds. It's called "overhead" and ranges from about half of research salary to double or more, depending on the institution. The fraction of overhead spent on journal subscription tends to be higher for smaller institutions.

I'm aware of overhead/indirects. They're not the only source of university funding, especially at public universities where you have tuition, course fees, a tax base, endowments, and a lot of other things, too. But, I'm not sure what you mean by saying it's half to double of research salary. I do know that overheads are typically at least 50% (and that's a very low bound for most institutions) of grant money received per grant.

Naturally smaller institutions are going to spend a larger piece of their pie on something if its price doesn't scale very much compared to a larger institution.

Well, when people say 50% overhead (the lower bound), they mean that the cost to the grant is `1.5*direct_cost`, not that half the grant goes to indirect costs.

Anyway, research is the primary motivation for purchasing journal subscriptions. It does not make financial sense for a non-research institution to subscribe to many journals. So regardless of which budget the subscriptions are formally taken from, they come from, it should be considered a cost of research. Making the system more efficient frees up money somewhere that reduces the cost of (externally- and internally-funded) research.

Maybe that's true of large universities, but I went to a school with around 3000 students, and most of the journal articles I looked for were not available. (At least in Computer Science.) Fortunately in CS you can track down the author's personal website, and they usually have all papers available there.